AGW Baseline Development

• AGW Baseline is a standardized framework that rigorously defines methodologies for constructing and validating reference protocols across finite fields, geophysics, computer vision, and astronomy.
• In finite fields, it provides explicit inversion frameworks for permutation polynomials using constant-time operations and precomputed small-table lookups to enhance cryptographic designs.
• In geophysical and imaging domains, the baseline offers validated models for acoustic-gravity wave propagation, ensemble-based statistical tests, and deep learning benchmarks for person re-identification and cosmological imaging.

The AGW baseline refers to rigorously defined reference frameworks or protocols constructed for methodology, benchmarking, or scientific measurement in fields where “AGW” indicates either the namesake of Akbary–Ghioca–Wang (in finite field theory) or “acoustic-gravity wave” (in geophysical/solar physics), or less commonly, other discipline-specific expansions (e.g., “Apparent Galaxy Wall” in astronomy, “AGW” model in computer vision). In each context, a baseline comprises a reproducible, validated foundation for construction, analysis, and inversion, typically emphasizing mathematical rigor, explicit algorithmic or modeling steps, and documented quantification of uncertainties.

1. AGW Baseline for Permutation Polynomials in Finite Fields

The Akbary–Ghioca–Wang criterion (AGW-criterion) provides a general construction for permutation polynomials (PPs) over finite fields and an explicit, compositional inversion framework. The AGW-baseline development in this context encompasses the following (Niu et al., 2020):

• Formalism: Given finite sets $A, S$ and maps $f: A \rightarrow A$, $g: S \rightarrow S$, along with surjective $\lambda, \mu: A \rightarrow S$ such that $\lambda \circ f = g \circ \mu$, $f$ is a bijection iff $g$ is a bijection and $f$ is injective on each fiber $\mu^{-1}(s)$ ($s \in S$) (AGW Lemma 2.1).
• Inversion Framework: General reduction of $f: A \rightarrow A$0 to inversion on a much smaller set via explicit bijections $f: A \rightarrow A$1 and an associated map $f: A \rightarrow A$2. The inverse is constructed as $f: A \rightarrow A$3 (Theorem 2.2).
• Algorithmic Baseline: For AGW-based PPs (multiplicative, additive, and hybrid forms), the inversion proceeds in constant-time field operations plus a single lookup or inversion in $f: A \rightarrow A$4, often feasible by precomputed small tables (see Table below).
• Cryptographic Baseline Integration: The AGW approach enables concise S-box or interleaver recipes with known and efficiently computable inverses, robust algebraic properties, and parameterization for side-channel and storage efficiency.

Construction Type	Inversion Formula (see (Niu et al., 2020))	Associated Reduced Map $f: A \rightarrow A$5
Multiplicative	$f: A \rightarrow A$6	$f: A \rightarrow A$7 on $f: A \rightarrow A$8
Additive	$f: A \rightarrow A$9	$g: S \rightarrow S$0 linear on $g: S \rightarrow S$1
Hybrid (I)	Explicit (Theorem 5.2)	$g: S \rightarrow S$2
Hybrid (II)	Explicit (Theorem 5.7)	$g: S \rightarrow S$3 on $g: S \rightarrow S$4

This methodology sets a reproducible algebraic and computational baseline for designing, inverting, and deploying AGW-based permutation structures in cryptography and coding (Niu et al., 2020).

2. AGW Baseline in Fluid and Solid Earth Physics: Acoustic-Gravity Waves

In geophysical/solar and ocean physics, the AGW baseline development comprises models and protocols for the generation, propagation, and detection of acoustic-gravity waves (AGWs) in stratified, compressible fluids. Prominent components include (Pethiyagoda et al., 27 Sep 2025, Fedorenko et al., 2020, Riashchikov et al., 2023):

• Canonical PDE Framework: Linearized compressible fluid equations under gravity, often in variable topography or stratification, with specified boundary/initial conditions.
• Dispersion Analysis: Rigorous identification of AGW branches (surface-gravity and high-n acoustic-gravity), eigenfunction expansion (modal decomposition), and parameters such as Brunt–Väisälä and acoustic cut-off frequencies.
• AGW Reconstruction Baseline: Protocols for reconstructing AGW parameters (e.g., neutral-density fluctuations, vertical displacements, periods) from indirect measurements such as VLF radio wave amplitude data, providing explicit conversion formulas between observables and AGW quantities. For example, the amplitude-based baseline relates measured VLF amplitude perturbation $g: S \rightarrow S$5 to $g: S \rightarrow S$6 with dimensionless scaling factor $g: S \rightarrow S$7, calibrated for each propagation path (Fedorenko et al., 2020).
• Numerical Implementation Baseline: Algorithmic steps for modal solution (LU-solve of block-sparse systems, Filon quadrature of frequency integrals, convergence with $g: S \rightarrow S$8 modes for pressure accuracy), standard benchmark scenarios (tsunami scenarios: shelf, trench, ridge), and reporting of pressure and surface elevation with physical parameterization (e.g., $g: S \rightarrow S$9, $\lambda, \mu: A \rightarrow S$0, $\lambda, \mu: A \rightarrow S$1, $\lambda, \mu: A \rightarrow S$2) (Pethiyagoda et al., 27 Sep 2025).
• Solar/Thermal Misbalance Baseline: For the solar atmosphere, explicit quartic dispersion relations including non-adiabatic (thermal misbalance) effects, introducing $\lambda, \mu: A \rightarrow S$3 as the governing regime parameter. The baseline encompasses analytical limits for cut-off and Brunt–Väisälä frequencies as functions of $\lambda, \mu: A \rightarrow S$4, describing transitions between adiabatic and thermally-dominated regimes (Riashchikov et al., 2023).

AGW Baseline	Key Equations / Protocols	Reference
Fluid AGW PDEs	$\lambda, \mu: A \rightarrow S$5; dispersion laws	(Pethiyagoda et al., 27 Sep 2025)
VLF/AGW Protocol	$\lambda, \mu: A \rightarrow S$6; stepwise inversion routine	(Fedorenko et al., 2020)
Solar Misbalance	$\lambda, \mu: A \rightarrow S$7	(Riashchikov et al., 2023)

These frameworks enable standardized simulation, measurement recovery, and theoretical analysis relevant to tsunami prediction, mesospheric diagnostics, and solar coronal heating.

3. AGW Baseline for Deep Learning and Computer Vision (Person Re-Identification)

In the domain of person re-identification (Re-ID), the “AGW baseline” denotes a composite state-of-the-art benchmark system comprising architectural, loss, and evaluation components (Ye et al., 2020):

• Architecture: ResNet-50 backbone (stride=1), five non-local attention blocks, generalized-mean (GeM) pooling, BNNeck, and a softmax classifier. Modal-specific “two-stream” variants address cross-modality (e.g., visible–infrared) tasks.
• Loss Function Baseline: Combination of identity loss (cross-entropy with label smoothing), center loss, and weighted regularization triplet loss (WRT) for hard positive/negative mining, with recommended weights $\lambda, \mu: A \rightarrow S$8.
• Training and Sampling Baseline: Batch-hard sampling (e.g., $\lambda, \mu: A \rightarrow S$9 for images), random erasing, warm-up and step-decay learning rate schedule, ImageNet normalization, Adam/SGD optimizers selected by modality.
• Evaluation Baseline: Introduction of mINP (mean Inverse Negative Penalty) metric for “hardest true match” retrieval—complementing standard rank-1 and mAP.
• Benchmark Performance: Demonstrated across four Re-ID subtasks (image, video, cross-modality, partial) with results competitive or superior to specialized SOTA models, and robust to domain shift, sampling changes, and hardware variability.

AGW Component	Details / Recipe	Source
Backbone	ResNet-50, stride=1, non-local blocks x5	(Ye et al., 2020)
Pooling	Generalized-mean pooling (GeM), $\lambda \circ f = g \circ \mu$0 per channel, learned	(Ye et al., 2020)
Losses	$\lambda \circ f = g \circ \mu$1	(Ye et al., 2020)
Evaluation	mINP: $\lambda \circ f = g \circ \mu$2	(Ye et al., 2020)

This AGW baseline delivers a reproducible, reference-grade framework for future Re-ID research, supporting transparent ablation, robust metrics, and adaptive deployment.

4. AGW Baseline in Astronomy: Apparent Galaxy Wall

Sailer defines the AGW “Apparent Galaxy Wall” baseline for deep field cosmological image analysis, establishing a quantifiable “visual coverage” effect threshold (Sailer, 2021):

• Formal Baseline: The AGW effect is declared when galaxy coverage $\lambda \circ f = g \circ \mu$3, where $\lambda \circ f = g \circ \mu$4 counts pixels above an extragalactic background light (EBL) threshold.
• Simulation Protocol Baseline: Both parameter-sensitivity and full-ensemble simulations are performed, with parameter grids across cosmological and image-formation uncertainties (e.g., $\lambda \circ f = g \circ \mu$5, $\lambda \circ f = g \circ \mu$6, $\lambda \circ f = g \circ \mu$7).
• Statistical Testing Baseline: A one-way one-sample t-test ($\lambda \circ f = g \circ \mu$8) is used to test whether the ensemble mean exceeds the 50% AGW threshold ($\lambda \circ f = g \circ \mu$9).
• Result Baseline: The 90-member ensemble yields $f$0; the effect is deemed “not likely” under mean uncertainties but not ruled out in edge cases.
• Recommendations: Ensemble analysis, sensitivity mapping, and future adjustment to AGW detection criteria with real JWST data (Sailer, 2021).

5. Implementation, Complexity, and Integration Aspects

AGW baseline protocols across these domains share rigorous algorithmic or mathematical construction, express complexity and storage bounds, and prioritize practical integration:

• Finite Fields: $f$1 algebraic steps and one small-table inversion per query ($f$2).
• Geophysical AGW: Full linear algebraic system solution, $f$3 for modal expansions; pressure fields converge with $f$4 error.
• Deep Learning: Standardized recipes for batch, schedule, optimizer, and ablation-centric verification.
• Astronomy: Explicit, ensemble-based simulation for robustness to parameter uncertainty.

Baseline protocols thus serve as both methodological minima and as reference points for further innovation, comparative study, or practical deployment.

6. Significance and Impact of AGW Baselines

Across its manifestations, the AGW baseline constitutes a foundational procedure by which the correctness, efficiency, reproducibility, and quantitative reliability of results in algebraic, physical, computational, and astronomical settings are assured. The AGW-criterion for finite fields enables efficient cryptosystem design with explicit inversion. Baselines for AGW propagation in oceans and atmospheres underpin both fundamental wave research and operational tsunami/mesosphere diagnostics. In computer vision, the AGW baseline system and metrics (mINP) structure the evaluation of Re-ID models for real-world surveillance and security use cases. In cosmology, clear AGW thresholds facilitate objective claims about field saturation and observational horizon limits.

All these baselines foster robust benchmarking, serve as “ground truths” for new theoretical and empirical innovations, and provide a reproducible starting point for domain-specific extensions and applications (Niu et al., 2020, Pethiyagoda et al., 27 Sep 2025, Fedorenko et al., 2020, Riashchikov et al., 2023, Ye et al., 2020, Sailer, 2021).